Biological Control 50 (2009) 129–136

Contents lists available at ScienceDirect

Biological Control

journal homepage: www.elsevier.com/locate/ybcon

Exclusion techniques reveal significant biotic mortality suffered by Asian psyllid (: ) populations in Florida citrus

Jawwad A. Qureshi *, Philip A. Stansly

University of Florida/IFAS, Department of Entomology and Nematology, Southwest Florida Research and Education Center, 2686 SR 29N, Immokalee, FL 34142, USA article info abstract

Article history: Management of ‘huanglongbing’ (HLB) or the Asian form of citrus greening disease caused by the bacte- Received 29 October 2008 rium, Candidatus Liberibacter asiaticus, and spread by Diaphorina citri Kuwayama depends largely on the Accepted 1 April 2009 management of the vector . Seventeen experiments conducted during 2006–2007 included D. citri Available online 8 April 2009 colonies initiated with eggs or first instar nymphs on young shoots protected with sleeve cages, sticky barriers or left unprotected through adult emergence to assess the relative impact of abiotic and biotic Keywords: factors on psyllid populations. Temperatures dropped to 0 °C and killed most of the shoots and nymphs Asian citrus psyllid in February 2007. Spiders (Araneae) and in the families Coccinellidae, Blattellidae, Chrysopidae, Citrus greening Formicidae, Syrphidae, Anthocoridae and Miridae were observed on colonies or caught in sticky barriers. Huanglongbing Integrated pest management The ladybeetles Olla v-nigrum, Curinus coeruleus, Harmonia axyridis, and Cycloneda sanguinea, the cock- Life tables roach Blattella asahinai, lacewings, Ceraeochrysa sp. and Chrysoperla sp., and spiders were most often Predation encountered. Parasitism by Tamarixia radiata provided 1–3% nymphal mortality in relation to adult emer-

Parasitism gence although a maximum of about 25% was observed. Net reproductive rate (R0), based on observed survivorship and temperature-dependent fecundity data was estimated to be 5- to 27-fold higher in the colonies protected with full cages compared to the unprotected colonies except during the freeze.

At the extreme, R0 was estimated in caged colonies at 125–285 in Jun 2006 whereas all nymphs disap- peared in the unprotected colonies. Intermediate results were observed in colonies protected with sticky barriers or coarse mesh cages. These findings suggest that biotic factors play a vital role in regulating pop- ulations of D. citri. Their elimination through reckless pesticide use could increase pest pressure and enhance the spread of the disease. Therefore, integrated control programs based on conservation of nat- ural enemies of D. citri through judicious use of insecticides and releases of new parasitoids are needed for sustainable management of pest and disease. Published by Elsevier Inc.

1. Introduction severely debilitating citrus trees (Aubert et al., 1996; Roistacher, 1996; Bové, 2006). In the United States, D. citri was first discovered The Asian citrus psyllid, Diaphorina citri Kuwayama, is one of in Palm Beach County, Florida on hedges of orange jasmine, Mur- the most important insect pests of citrus in several tropical and raya paniculata (L.) Jack. (Rutaceae) in 1998 (Halbert, 1998). It is subtropical regions of the world, and is also reported from related now well established in citrus growing regions of Florida (Qureshi plant species in family Rutaceae (Halbert and Manjunath, 2004). et al., 2009), Texas (French et al., 2001), and has most recently been Shoots with unfolded and tender leaves are needed for oviposition detected in California (http://www.cdfa.ca.gov/egov/Press_Re- and nymphal development; whereas adults can survive, but not leases/Press_Release.asp?PRnum=08-057). Citrus greening disease reproduce, on mature leaves (Shivankar et al., 2000; Michaud, was first identified from citrus groves in south Florida during 2004; Qureshi and Stansly, 2008). Nymphs and adults directly 2005 (Halbert, 2005) and is spreading rapidly throughout the state damage the plants by feeding on the young shoots and leaves, (FDACS-DPI, 2008). although the primary importance of D. citri is as a vector of the bac- The development of D. citri is dependent upon temperature, terium, Candidatus Liberibacter asiaticus, one of the causal organ- with an optimum range of 25–28 °C(Liu and Tsai, 2000). Under isms of Huanglongbing (HLB) or citrus greening disease (Garnier controlled conditions, populations reared on M. paniculata at et al., 2000; Bové, 2006). Citrus greening is considered the most 10 °C and 33 °C failed to develop, whereas survival of the nymphal important disease of citrus, reducing fruit yield and quality and instars 3–5 was unchanged between 15 and 28 °C. Highest intrinsic rate of increase (0.1999) and net reproductive rates (292.2) were * Corresponding author. Fax: +1 239 658 3469. observed at 28 °C. Fecundity increased with increasing tempera- E-mail address: jawwadq@ufl.edu (J.A. Qureshi). ture, reaching a life time maximum of 748 eggs per female at

1049-9644/$ - see front matter Published by Elsevier Inc. doi:10.1016/j.biocontrol.2009.04.001 130 J.A. Qureshi, P.A. Stansly / Biological Control 50 (2009) 129–136

28 °C. Mean developmental period from egg to adult varied from study. Blocks containing experimental trees were planted at the 49.3 days at 15 °C to 14.1 days at 28 °C(Liu and Tsai, 2000). rate of 323 trees ha1 on double-row raised beds, irrigated by mi- Biological and chemical methods of pest control and removal of cro-sprinklers and otherwise subjected to conventional cultural disease-infected trees are tactics being employed to reduce the practices (Jackson, 1999). Climatic data were obtained from a incidence of the pest and the disease in Florida (Michaud, 2004; weather station located within 500 m of the study blocks Qureshi and Stansly, 2007; Rogers et al., 2008). Biological control (www.fawn.ifas.ufl.edu). Minimum, maximum, and mean daily has been an important component of citrus insect pest manage- temperature averaged 9.5–22.7 °C, 23.7–33.6 °C, and 16.5– ment in Florida (McCoy, 1985) and therefore a thorough under- 26.8 °C, respectively, and relative humidity averaged 66–84% dur- standing of its contribution to mortality of D. citri is essential ing the study period (Fig. 1). (Stansly and Qureshi, 2008). Although natural enemies were not seen as key factors in regulating populations of D. citri in an early 2.2. Cohort establishment and experimental procedures study conducted in southern Florida during 1998–1999 (Tsai et al., 2002), Michaud (2001) reported increased populations of a Experiments were initiated on 23 Jan, 13 Mar, 24 Apr, 8 May, 16 native predatory ladybeetle, Olla v-nigrum Mulsant, in response Jun, 3 and 31 Jul, 21 Aug, 18 Sept, 16 Oct, 13 Nov, and 4 Dec in 2006, to the D. citri invasion. He found coccinellid beetles, Harmonia axy- and 5 Jan, 13 Feb, 13 Mar, 16 Apr, and 11 May in 2007. The first three ridis Pallas, O. v-nigrum, Cycloneda sanguinea L., and Exochomus experiments in 2006 were conducted on 10 year old trees and the childreni Mulsant as the most important predators of high-density remainder on 5–6 year old trees. We use the term ‘cohort’ for the nymphal populations of D. citri, followed by lacewings, Ceraeo- synchronous age-class of insects in the grove and ‘colony’ for chrysa sp. and Chrysoperla rufilabris Burmeister, and the spider nymphs of uniform age cohabiting a single shoot. Each experiment (Araneae), Hibana velox (Becker) during the course of three cohort began with 20–40 young shoots infested with a colony of D. citri studies conducted Sept–Oct 2001 and Jun–Jul 2002 and surveys of eggs or neonates that were examined at least three times a week three groves in Lake Alfred, FL and one in Ft. Pierce, FL during Feb– until no nymphs were left. Shoots were infested with feral popula- Mar 2002 (Michaud, 2004). In a two year study in southwest Flor- tions except when insufficient infested shoots were available for ida, Curinus coeruleus Mulsant, H. axyridis, O. v-nigrum, C. sanguinea experiments initiated on 13 Nov and 4 Dec 2006 and 5 Jan 2007. L., lacewings and spiders were found in orange trees infested with At those times, gravid females from greenhouse colonies were D. citri (Qureshi and Stansly, 2008). caged on the shoots for 24–48 hr to obtain eggs. Infested shoots Tsai et al. (2002) and Michaud (2004) reported only 1–2% para- were selected on multiple trees and covered individually with sitism by Tamarixia radiata (Waterston), an ectoparasitoid of D. citri 15.24 30.48 cm (6 12 in.) sleeve cages made from fine mesh or- nymphs imported from Taiwan and south Vietnam, and released in gandy for 24–48 h to allow for sufficient expansion of tightly ap- Florida to improve the biological control of D. citri (Hoy et al., pressed leaves in the bud in order to observe all eggs and 1999). Parasitism rates of T. radiata calculated based on adult nymphs. Cages were removed and all eggs and neonate nymphs emergence of pest and parasitoid were found to be variable in a were counted on each shoot. Older instars were either removed study conducted in 28 citrus groves across Florida during 2006– or shoots containing them were not selected. Selected shoots were 2007, averaging less than 20% during spring and summer but tagged with flagging tape, numbered, and randomly distributed increasing during fall to 39% in Sept. and 56% in Nov in central across one to three exclusion treatments and an unprotected con- and southwest Florida, respectively (Qureshi et al., 2009). The trol. Ten infested shoots were allocated to each treatment. entomopathogenic fungus, Hirsutella near citriformis Speare, is also Three exclusion techniques or treatments were used to evaluate known to attack D. citri in Florida; however, its contribution is not the contribution of natural factors, particularly predators and par- well documented (Meyer et al., 2007). asitoids to the mortality of D. citri populations. Sleeve cages as de- Observations of natural enemy impact and the need to develop scribed above were used to completely protect against predators sustainable pest and disease management programs warranted and parasitoids. The same size cage made with coarse (1 mm2) more detailed investigations to estimate the contribution of natural mesh organdy was employed on 10 colonies in experiments initi- mortality factors to population growth of D. citri in the field. Evalu- ated 8 May, 16 Jun, 3 Jul, 31 Jul, and 6 Dec in 2006 and all experi- ations through direct observation of insect colonies established un- ments in 2007 to allow T. radiata individuals, but not larger insects der different conditions have proven to be useful in similar to enter the cage. A third exclusion treatment, included from 13 investigations (Southwood, 1978; Elkington et al., 1992; Van Drie- Mar 2006 onward, consisted of a sticky barrier (Tangle footÒ insect sche et al., 1994; Naranjo and Ellsworth, 2005; Xiao et al., 2007). trap coating, The Tanglefoot Company, Grand Rapids, Michigan

Net reproductive rate (R0) or the lifetime reproductive output of a 49504) applied to a five centimeter portion of the branch to stop single female is a population parameter often used to analyze the crawling insects from reaching the psyllid colonies. The colonies impact of mortality in cohorts (Birch, 1948; Bellows and Van Drie- used as control were without any exclusion barrier of any kind sche, 1999; Carey, 2001). The objectives of our studies were to esti- (Smith and DeBach, 1942). mate the overall contribution of abiotic and biotic factors to natural mortality of D. citri cohorts or generations at different times and simultaneously identify the sources related to such mortality.

2. Materials and methods

2.1. Location and experimental conditions

All experiments were conduced in a 28 ha experimental citrus orchard at the Southwest Florida Research and Education Center, University of Florida – IFAS, Immokalee, FL (26°2800N, 81°2600W) during 2006–2007. The experimental trees were five to ten year old sweet orange Citrus sinensis (L.) Osbeck ‘Valencia’ and were Fig. 1. Minimum, maximum and average temperature (°C) and relative humidity not treated with insecticides before or during the course of the (%) during experiments initiated on different dates, 2006–2007. J.A. Qureshi, P.A. Stansly / Biological Control 50 (2009) 129–136 131

Shoots were examined with a 10 hand lens Mondays, Wednes- cages, most notably in 2006 (Fig. 2). Differences during 2007 were days and Fridays except on holidays. All eggs, nymphs, and any less pronounced, and non-existent during the February cohort emerging adults or their exuviae were counted on each observation. when most flushes and nymphs were killed by a freeze during Adults, exuviae, and any additional eggs were removed following which temperatures dipped to 0 °C for two successive nights dur- each observation. Mortality was estimated directly by the disap- ing the experiment. Most loss in unprotected colonies resulted pearance of nymphs or indirectly by their emergence as adults, from the disappearance of instars 1–3 which took about 12 days and considered as generational mortality (Naranjo and Ellsworth, to develop (Fig. 2). Disappearance of these nymphs by day 12 2005). Any predators or parasitoids observed on the shoots or sticky was more marked (76–100%) between Jan and Oct 2006 than barrier were counted and removed from the latter. Additionally Nov–Dec (29–31%). More nymphs were lost by day 12 during predator populations were assessed beginning 8 May 2006 on each 2006 (76–92% Jan–May) than the same period in 2007 (36–76%) experimental tree by searching the foliage for 1 min at approach- when insecticide use on citrus had accelerated in the region. Few able heights around the canopy. nymphs were found trapped in the sticky barrier indicating that Overall rates of parasitism by T. radiata within study groves losses were not due to dispersal (data not presented). were assessed by collecting 4th and 5th instar D. citri nymphs that were reared through to emergence of adult psyllids or parasitoids 3.2. Population parameters in the laboratory. Shoots containing the nymphs were held in glass cylinders with the stems in water and covered with ventilated con- Net reproductive rate (R0), estimated for colonies protected tainers held at ambient temperature in the laboratory for at least with a full cage, averaged 191.1 ± 13.2 and ranged from a low of 2 weeks to allow adults of D. citri and T. radiata to emerge. 73.7 in Jan 2006 to a high of 285.2 in Jun 2006, discounting the fro-

zen cohort Feb 2007 (Table 1). In comparison, R0 in unprotected 2.3. Statistical analysis colonies averaged 28.1 ± 5.2, ranging from a low of zero in Jun

2006 to a high of 64.3 in Mar 2006 and Apr 2007. R0 was signifi- Survival for each colony was calculated as: (no. adults emerged/ cantly lower for all unprotected colonies compared to full cage col- max. no. immatures observed) 100. The net reproductive rate (R0) onies except Feb 2007 during the freeze (Table 1). Differences for was calculated for each colony as the product of nymphal survival to the remaining colonies averaged 163 ± 15.3 with highest 285.2 in adulthood and fecundity (Bellows and Van Driesche, 1999; Xiao Jun 2006 when there was no survival in unprotected colonies et al., 2007) assuming a sex ratio of 1:1 using fecundity at the appro- and 161.3 in May 2007. priate temperature range as reported by Liu and Tsai (2000). Mean Net reproductive rate in colonies protected by sticky barriers generation time (T) was estimated as the sum of egg incubation time and partial cages were generally intermediate between unpro- (Liu and Tsai, 2000) and nymphal developmental time through 50% tected and full caged colonies (Table 1). Differences between sticky adult emergence as observed in each colony, plus the pre-reproduc- barrier and uncaged colonies were significant for 8 out of 16 exper- tive period as reported by Wenninger and Hall (2007). The rate of iments and between sticky barrier and full cage colonies for 13 out nymphal development to adulthood for colonies that developed in of 16 experiments. Similarly, 5 out of 10 and 6 out of 10 compari- full cages and were therefore not influenced by predators or parasit- sons between partial and full cage colonies and between partial oids was calculated as the reciprocal of days to 50% adult emergence, and no cage colonies, respectively, were significant. Four of 10 and analyzed for relationship with temperature. comparisons, all in 2006, between partial cage and sticky barrier Numbers of immature and adult predators were combined to exclusion were significant with higher R0 in the partial cage, dem- compare relative abundance of different groups (ladybeetles, lace- onstrating more protection with this method. wings, and spiders) and ladybeetles species (C. coeruleus, H. axyri- The time required for 50% adult emergence varied with average dis, O. v-nigrum, and C. sanguinea). Data were transformed by temperature (R2 = 0.71; F = 35.85; df = 1, 16; P < 0.0001), resulting log(x + 1) for analysis. Parasitism rates (%) from field collected lab- in short generation times during warm seasons. Generation times oratory reared nymphs were calculated as: (no. adults emerged T. were shortest (mean ± SEM = 18.8 ± 0.5 days) in Aug 2006 and lon- radiata/total no. adults emerged D. citri + T. radiata) 100 (Stansly gest (mean ± SEM = 39.03 ± 1.6 days) in Nov–Dec 2006 excluding et al., 1997; Qureshi et al., 2009). A comparable index was calcu- the Feb 2007 cohort (H = 121.57; df = 16; P < 0.0001). Caged colo- lated from the number of mummies and adult psyllids developing nies consistently took more time to terminate than unprotected in colonies in the field. colonies (Fig. 2) despite expected higher temperature in the cages, Data were analyzed for normality using the Shaprio-Wilk W test probably because fewer individuals survived to emergence in and normality plots using the univariate procedure (Shapiro and unprotected cohorts. Wilk, 1965; Shapiro et al., 1968; SAS Institute, 2004). Data with nor- mal distribution were analyzed using analysis of variance (ANOVA) 3.3. Natural enemies and GLM procedure (SAS Institute, 2004). Predator data collected over time on the same trees during a cohort were analyzed using re- 3.3.1. Predators peated measures ANOVA and PROC MIXED procedure (SAS Insti- Spiders and insect predators in the families Blattellidae, Cocci- tute, 2004). Treatment means were compared using the least nellidae, Chrysopidae, Formicidae, Syrphidae, Anthocoridae and significant difference (LSD) at a significance level of 0.05 (Littell Miridae were observed on colonies or trapped in sticky barriers et al., 1996). Non-normal data were analyzed by using the non- (Fig. 3). The Asian cockroach Blattella asahinai Mizukubo was the parametric Kruskal–Wallis test and pairwise treatment compari- most often encountered species trapped in sticky barriers but sons made with the Mann–Whitney U-test (Hollander and Wolfe, never seen in colonies, presumably because of its nocturnal habits 1973; SPSS, 2004). Actual means ± SEM are presented for all data. and residence in the litter during the day. Spiders were observed with about equal frequency trapped in sticky barriers or on colo- 3. Results nies, whether uncaged or protected by barriers, which they pre- sumably accessed by dropping from above or bridging. The 3.1. Cohort survival lacewings Ceraeochrysa sp. and Chrysoperla sp. were most often seen as adults or larvae in control colonies or adults in sticky bar- Young nymphs generally disappeared rapidly from unprotected riers, indicating that adults both walked and flew to psyllid-inhab- colonies causing rapid divergence from the situation in small mesh ited flush. Ants were only seen in unprotected colonies or trapped 132 J.A. Qureshi, P.A. Stansly / Biological Control 50 (2009) 129–136

Fig. 2. Mean number of Diaphorina citri nymphs remaining through adult emergence in colonies established on shoots of Citrus sinensis that were fully caged or not caged (no exclusion barrier) to assess the impact of natural mortality factors on psyllid populations, 2006–2007. Dates in the figures indicate initiation of the experiment.

in sticky barriers which were effective in excluding them. Ladybee- dis (Fig. 4B). In spring 2007, O. v-nigrum was the only commonly tles comprised of the species O. v-nigrum, C. sanguinea, C. coeruleus, observed species, with others present at low levels. and H. axyridis were the most frequently encountered group. Lady- beetle larvae were seen in control colonies and trapped in sticky 3.3.2. Parasitoids barriers, but not in colonies protected by sticky barriers, indicating Tamarixia radiata was the only parasitoid that emerged from D. that they were unable to transverse this impediment. citri nymphs. Parasitism in colonies was observed during Nov–Dec The relative number of adults captured in sticky barriers was 2006 and Apr–May 2007, although a mere 1–3% of all nymphs greatest for C. sanguine, the species least well adapted to a psyllid were parasitized in any one experiment. This corresponded to 2– diet (Michaud and Olsen, 2004). In contrast, O. v-nigrum, known as 24% when considered relative to adult emergence. The greatest a psyllid specialist, was most predominantly observed as larvae in share of parasitism (%) occurred in colonies protected by sticky control colonies. More ladybeetle adults were trapped in sticky barriers, averaging 9.5 ± 9.5 and 4.3 ± 4.3 in Nov and Dec 2006 barriers than adults or larvae observed on colonies except O. v-ni- and 15. 9 ± 8.2 and 24.4 ± 8.7 in Apr and May 2007. In contrast, par- grum for which more larvae were seen on the colonies (Fig. 3). asitoids accounted for 2.5–6.3% emergence from unprotected colo- Ladybeetles were more abundant than lacewings and spiders nies in Dec, Apr and May with none seen in Nov. Parasitized among the three commonly observed groups during the 1 min vi- nymphs were observed in only one partial cage during Dec, sual search of tree canopy except in Aug and Dec 2006 when no dif- accounting for 2.5 ± 2.3% of total emergence. Parasitism rates from ferences were observed (P = 0.05, Fig. 4A). Numbers of ladybeetles field collected laboratory reared nymphs ranged between 0% and declined from May through Aug and only a few were seen the rest 14% (average = 4%) in spring and summer and 0–45% (aver- of the year. In 2006, O. v-nigrum and C. coeruleus were the most age = 26%) in fall (Table 2). In 2006, parasitism declined from 14% commonly observed species followed by C. sanguinea and H. axyri- in May to 2% in Jul and was not observed again until fall when J.A. Qureshi, P.A. Stansly / Biological Control 50 (2009) 129–136 133

Table 1 Net reproductive rate (Mean ± SEM) of Diaphorina citri calculated based on fecundity (Liu and Tsai, 2000) and survival of nymphs in colonies that were unprotected (no exclusion barrier) or protected with full cages, partial cages and sticky barriers to assess the impact of natural mortality factors on psyllid populations, 2006–2007.

Cohort initiation datea No exclusion Full cage exclusion Partial cage exclusion Sticky barrier exclusion F or HP 2006 23-Jan 22.57 ± 7.53b 73.71 ± 19.20a – – 4.55 0.0329 13-Mar 64.27 ± 20.00b 169.84 ± 22.54a – 125.83 ± 26.53a 7.12 0.0053 24-Apr 25.66 ± 12.39b 170.87 ± 38.43a – 116.59 ± 23.80a 10.63 0.0009 8-May 52.66 ± 31.14c 239.23 ± 10.44a 188.16 ± 19.05b 62.78 ± 18.88c 22.9 <0.0001 16-Jun 0.00d 285.20 ± 13.84a 205.85 ± 32.59b 125.28 ± 31.83c 25.22 <0.0001 3-Jul 37.17 ± 13.87c 236.29 ± 10.34a 158.89 ± 26.04b 68.38 ± 16.43c 25.37 <0.0001 31-Jul 6.41 ± 4.35c 244.16 ± 26.57a 165.40 ± 34.78ab 135.01 ± 44.06b 10.27 0.0001 21-Aug 25.91 ± 11.49b 174.53 ± 27.90a – 61.94 ± 24.16b 12.58 0.0004 18-Sep 4.47 ± 4.47c 261.75 ± 21.80a – 105.60 ± 29.99b 30.87 <0.0001 16-Oct 14.00 ± 10.27c 170.46 ± 18.11a – 96.99 ± 20.78b 25.32 <0.0001 13-Nov 13.24 ± 8.54c 194.07 ± 16.46a – 84.01 ± 28.79b 22.58 <0.0001 4-Dec 33.05 ± 19.30b 176.53 ± 21.24a 139.14 ± 21.30a 37.87 ± 13.97b 16.88 <0.0001 2007 5-Jan 50.74 ± 20.17b 190.15 ± 25.38a 105.16 ± 24.99b 45.43 ± 16.27b 9.3 0.0002 13-Feb 8.65 ± 8.65a 13.59 ± 9.47a 38.70 ± 20.32a 35.43 ± 13.19a 1.32 0.2868 13-Mar 7.89 ± 4.09c 131.46 ± 26.89a 65.06 ± 22.77bc 68.90 ± 23.89b 6.21 0.0024 16-Apr 64.34 ± 20.80b 150.83 ± 18.49a 99.90 ± 29.36ab 66.43 ± 18.41b 3.36 0.0334 11-May 27.64 ± 8.74b 188.94 ± 26.76a 51.13 ± 14.81b 45.30 ± 14.70b 15.35 0.0015

a Data for colonies initiated on 8 and 11 May and 3 Jul analyzed using Kruskal–Wallis test followed by Mann–Whitney U test, all others using normal ANOVA followed by LSD test. Means within a row followed by a different letter are significantly different (P = 0.05). –, no treatment.

Fig. 3. Total number of predators observed on colonies of Diaphorina citri nymphs on Citrus sinensis shoots or trapped in the sticky barriers used to protect colonies initiated on different dates, 2006–2007. Note: Blattella asahinai and Araneae data presented as adults include some juveniles not recorded as separate age class.

26–45% was observed in Nov–Dec (Table 2). This was followed by a in the absence of mortality factors impeded by the cage. Sticky bar- decrease to 4% in Jan 2007 and a slight increase to 9% in May. riers reduced R0 an average 3-, 12- and 5-fold between Jan–May, Jul–Sept, and Oct–Dec of 2006 and 3-fold between Jan and May 4. Discussion of 2007 compared to the unprotected colonies. This constituted an approximately 50% reduction compared to populations pro- Impressive differences were observed in survival and conse- tected with complete cages and was attributable to predators that quently in estimated net reproductive rate between unprotected were observed on the colonies or caught in the sticky barrier colonies compared to colonies protected with fine mesh cages. (Fig. 3). The partial cage treatment was introduced to admit para- Sticky barriers and coarse mesh cages provided partial protection. sitoids but exclude predators. However, parasitism was negligible These results suggested a leading role for predation among mortal- in control colonies and partial cages. Nevertheless, estimated R0 ity factors impacting populations of D. citri (Table 1). Estimated R0 was 4-fold greater in coarse mesh cages than unprotected colonies for completely protected colonies compared to unprotected colo- in May, Jul, and Dec 2006 and Jan–May 2007, and 26-fold greater in nies was greater by an average 5-, 27-, 11-, and 6-fold during Aug 2006. The cause was mortality inflicted by small predators Jan–May, Jul–Sept, Oct–Dec in 2006 and Jan–May of 2007, respec- that penetrated the coarse mesh and were seen on the colonies tively. Thus, D. citri showed great potential for population increase such as early instar ladybeetle larvae and small spiders. Increased 134 J.A. Qureshi, P.A. Stansly / Biological Control 50 (2009) 129–136

served nymphs and adults of D. citri developing and surviving at temperatures between 35 and 40 °C in the greenhouse (J.A. Qure- shi, personal observations). In contrast, different levels of protec- tion resulted in great differences in survivorship, indicating that most mortality was biotic. Ladybeetles were the most frequently observed visitors to colo- nies although spiders, cockroaches, lacewings, ants, syrphids, min- ute pirate bugs and mirids were also seen in the colonies or caught in sticky barriers. Ladybeetles were most common in the spring, declining during summer, and largely absent during winter (Fig. 4) in concert with flushing patterns and psyllid abundance (Qureshi et al., 2009). A reduction in ladybeetles and consequently psyllid mortality observed in 2007 compared to the previous year may have been due to the increased foliar applications of insecti- cides against D. citri that negatively impact beneficial insects (Rog- ers, 2008). The four ladybeetle species observed, O. v-nigrum, C. sanguinea, C. coeruleus and H. axyridis, are well documented predators of psyl- lid immatures. Michaud and Olsen (2004) showed that O. v-nigrum, C. coeruleus and H. axyridis developed and reproduced successfully on D. citri nymphs obtained from colonies on M. paniculata, whereas, C. sanguinea developed successfully but could not repro- duce as adults. The greater abundance of O. v-nigrum compared to the other three species observed through presence of its larvae on the unprotected colonies and its populations in both years suggest Fig. 4. Mean (±SEM) numbers of spiders, lacewings and ladybeetles per 1 min dominance among Coccinellidae as predators of D. citri immatures. observation of foliage around Citrus sinensis trees that contained colonies of Survival of psyllid nymphs was reduced to zero in Jun 2006 when Diaphorina citri nymphs initiated on different dates, 2006–2007. O. v-nigrum was most abundant (Fig. 4). Populations of this species were reported to increase in response D. citri invasion in Florida (Michaud, 2002). More C. sanguinea adults trapped in the sticky Table 2 Percentage parasitism by Tamarixia radiata from 4th and 5th instars nymphs of barriers than seen on the colonies indicated that they may walk Diaphorina citri collected from Citrus sinensis trees in the experimental blocks and away from infested shoots more often than other species. Large dif- reared in the laboratory through adult emergence, 2006–2007. ferences in R0 between unprotected and protected colonies were Collection date D. citri nymphs Adults emerged Parasitisma (%) observed in summer, and were probably due to combined efforts instars 4–5 of ladybeetles with other predators which were common late in D. citri T. radiata Total the season. Sticky barriers provided significant protection during 5/15/2006 238 31 5 36 13.89 this period, confirming that crawling predators were still active 5/19/2006 170 37 3 40 7.50 6/21/2006 120 47 2 49 4.08 and likely responsible for the differences noted. 7/10/2006 92 59 1 60 1.67 Spiders were another dominant group of predators observed in 8/9/2006 120 27 0 27 0.00 equal numbers on colonies and sticky barriers, although infre- 8/23/2006 300 60 0 60 0.00 quently noted by visual observation of trees due to their cryptic 9/18/2006 41 16 0 16 0.00 habits. Spiders were seen in equal numbers on unprotected colo- 11/14/2006 50 17 14 31 45.16 11/20/2006 68 17 6 23 26.09 nies and colonies protected with sticky barriers, indicating their ar- 12/1/2006 150 23 0 23 0.00 rival by dropping down, ballooning or bridging. Spiders were 12/20/2006 482 88 44 132 33.33 reported to feed on citrus psyllids by Shivankar et al. (2000) and 1/10/2007 287 106 4 110 3.64 Michaud (2002, 2004) signaled salticids and anyphaenids as possi- 3/13/2007 192 42 0 42 0.00 4/16/2007 199 36 3 39 7.69 bly important predators of immature D. citri. Eighteen species of 5/19/2007 215 31 3 34 8.82 spiders, mostly from the family Salticidae, were identified as pre- dators of Trioza erytreae (DelGuercio) also known as African citrus a Parasitism (%) calculated as the ratio of number of emerged adult T. radiata to the total combined number of adult D. citri and T. radiata that emerged from psyllid (Hemiptera: Triozidae) (Van den Berg et al., 1992). Reduced samples of 4th and 5th instars nymphs (Stansly et al., 1997; Qureshi et al., 2009). survival in the unprotected colonies at times when ladybeetles and lacewings were relatively scarce indicated that other predators such as spiders inflicted significant mortality. The Asian cockroach, an introduced insect detected in Florida in

R0 seen in response to partial or complete exclusion reflected 1986 and established in Georgia, Alabama, and Texas (Snoddy and the action of some or all components of the predator complex on Appel, 2008; Pfannenstiel et al., 2008) was the most frequently D. citri populations. trapped insect in sticky barriers. The cockroaches live in the litter Temperature drove nymphal development rate but had little ef- and move into the canopy at night where they may feed on young fect on mortality except for the freezing event in Feb 2007. Average flush (Stansly et al., 1999). Asian cockroach has been reported to temperatures during the experiments were within the range found prey on brown citrus aphids in citrus (Persad and Hoy, 2004) and suitable for survival and development of D. citri in the laboratory eggs of lepidopteran pests in cotton and soybean in Texas (Pfann- (Liu and Tsai, 2000). These authors reported a decline in the survi- enstiel et al., 2008). We observed very little impact on eggs and vorship of young instars at temperatures above 28 °C. Very few early instar nymphs of D. citri in cage studies (J.A. Qureshi, unpub- dead nymphs were observed between Jun and Sept when maxi- lished data) but further investigations are in progress. Other preda- mum temperatures reached 32–33 °C, although some dead tors in the families Syrphidae, Anthocoridae and Miridae were seen nymphs may have fallen off the shoots. Furthermore, we have ob- in very small numbers on colonies or in sticky barriers. J.A. Qureshi, P.A. Stansly / Biological Control 50 (2009) 129–136 135

The overall impact of parasitism by T. radiata was a mere 1–3% Bellows, T.S., Van Driesche, R.G., 1999. Life table construction and analysis for of the total mortality and therefore had a negligible effect on net evaluating biological control agents. In: Bellows, T.S., Fisher, T.W. (Eds.), Handbook of Biological Control; Principles and Applications of Biological reproductive rate. Parasitism was most frequently observed in col- Control. Academic Press, NY, p. 1046. onies protected by sticky barriers, accounting for 24% of all adults Birch, L.C., 1948. The intrinsic rate of natural increase of an insect population. emerged in May 2007. These results mirrored those of Michaud Journal of Ecology 17, 15–26. Bové, J.M., 2006. Huanglongbing: a destructive, newly-emerging, century-old (2004) who observed less than 2% mortality from parasitism in disease of citrus. Journal of Plant Pathology 88, 7–37. unprotected colonies, but 20 times more in colonies protected Carey, J.R., 2001. Insect biodemography. Annual Review of Entomology 46, 79–110. from predators by open (2 mm) mesh cages, although overall mor- Elkington, J.S., Buonaggorsi, J.P., Bellows, T.S., Van Driesche, R.G., 1992. Marginal attack rate, k-values and density dependence in the analysis of tality was less than in unprotected colonies. We used a smaller contemporaneous mortality factors. Researches on Population Ecology 34, 29– (1 mm) mesh that was penetrated only once by ovipositing T. rad- 44. iata. However, the sticky barrier also appeared to provide partial (FDACS-DPI) Florida Department of Agriculture and Consumer Services, Division of Plant Industries, 2008. Huanglongbing (HLB)/Citrus Greening Disease. Available protection from intraguild predation with a similar increase in par- from: http://www.doacs.state.fl.us/pi/chrp/greening/citrusgreening.html. asitism but decrease in overall mortality. Nevertheless, parasitism French, J.V., Kahlke, C.J., da Graça, J.V., 2001. First record of the Asian citrus psylla, rates seen from nymphs samples collected from the experimental Diaphorina citri Kuwayama (Homoptera: Psyllidae), in Texas. Subtropical Plant blocks were generally higher than even colonies partially protected Science 53, 14–15. Garnier, M., Jagoueix-Eveillard, S., Cornje, H.F., Le Roux, P.R., Bové, J.M., 2000. by sticky barriers, indicating that intraguild predation in general Genomic characterization of a Liberibacter present in an ornamental rutaceous may be less prevalent than observed in the colonies. Parasitism tree, Calodendrum capense, in the Western Cape province of South Africa. rates from reared samples agreed with results from a state-wide Proposal of ‘Candidatus Liberibacter africanus subsp. capensis’. International Journal of Systematic and Evolutionary Microbiology 50, 2119–2125. study showing that parasitism rates were variable and generally Halbert, S.E., 1998. Entomology Section. Triology 37, 6–7. low (<20%) during spring and summer but increased to 39–56% Halbert, S.E., 2005. Pest Alert: Citrus Greening/Huanglongbing. Florida Department in fall (Qureshi et al., 2009). This finding warranted ongoing foreign of Agriculture and Consumer Services, Division of Plant Industry. Available from: http://www.doacs.state.fl.us/pi/chrp/greening/citrusgreeningalert.html. exploration and possible augmentative releases of D. citri Halbert, S.E., Manjunath, K.L., 2004. Asian citrus psyllids (: Psyllidae) parasitoids. and greening disease of citrus: a literature review and assessment of risk in Our results suggest that natural mortality factors, particularly Florida. Florida Entomologist 87, 330–353. Hall, D.H., 2008. Biological control of Diaphorina citri. In: Proceedings of the predators and to lesser extent parasitoids, impose significant mor- International Workshop on Huanglongbing of Citrus (Candidatus Liberibacter) tality on populations of D. citri. The estimated net reproductive and the Asian Citrus Psyllid (Diaphorina citri). Available from: http:// rates based on fecundity data from a laboratory study (Liu and Tsai, 148.243.71.63/includes/asp/download.asp?iddocumento=1746&idurl=3583. Hollander, M., Wolfe, D.A., 1973. Nonparametric Statistical Methods. John Wiley & 2000) are probably over-estimated given that adult survivorship Sons Inc., New York, NY. and thus fecundity is likely to be lower in the field. Inspite of biotic Hoy, M.A., Nguyen, R., Jeyaprakash, A., 1999. Classical biological control of the Asian mortality, Surviving populations have proved sufficient to cause citrus psylla-release of Tamarixia radiata. Citrus Industry 80, 20–22. economically significant disease transmission. On the other hand, Jackson, L.K., 1999. Citrus Growing in Florida. Univ. Florida Press, Gainesville. p. 408. Littell, R.C., Milliken, G.A., Stroup, W.W., Wolfinger, R.D., 1996. SAS System for elimination of biotic mortality by intensive use of insecticides Mixed Models. SAS Institute, Cary, NC. could require an impractical level of efficiency against increased Liu, Y.H., Tsai, J.H., 2000. Effects of temperature on biology and life table parameters incidence of pest and disease that might not be cost effective and of the Asian citrus psyllid, Diaphorina citri Kuwayama (Homoptera: Psyllidae). Annals of Applied Biology 137, 201–216. would likely be inductive of secondary pest outbreaks. A preferable McCoy, C.W., 1985. Citrus: current status of biological control in Florida. In: Hoy, alternative is conservation and enhancement of biotic mortality M.A., Herzog, D.C. (Eds.), Biological Control in Agricultural IPM Systems. through the use of selective insecticides, application methods Academic, Orlando, FL, pp. 481–499. Meyer, J.M., Hoy, M.A., Boucias, D.G., 2007. Morphological and molecular and timings, coupled with inoculative and/or augmentative release characterization of a Hirsutella species infecting the Asian citrus psyllid, of natural enemies. Examples of selective insecticide use include Diaphorina citri Kuwayama (Hemiptera: Psyllidae), in Florida. Journal of (1) soil-applied systemic insecticides that avoid direct contact with Invertebrate Pathology 95, 101–109. Michaud, J.P., 2001. Numerical response of Olla v-nigrum (Mulsant) (Coleoptera: natural enemies (Qureshi and Stansly, 2007, 2008), (2) dormant Coccinellidae) to infestations of Asian citrus psyllid (Hemiptera: Psyllidae) in sprays directed at adults when trees are not flushing and predators Florida. Florida Entomologist 84, 608–612. largely absent (Stansly and Qureshi, 2008), and (3) selective insec- Michaud, J.P., 2002. Biological control of Asian citrus psyllid (Homoptera: Psyllidae) in Florida. A preliminary report. Entomological News 113, 216–222. ticides such as horticultural mineral oils, lipid synthesis inhibitors Michaud, J.P., 2004. Natural mortality of Asian citrus psyllid (Homoptera: Psyllidae) and spinosyns (Stansly et al., 2002). At the same time, biological in central Florida. Biological Control 29, 260–269. control might be enhanced by inoculative releases of new strains Michaud, J.P., Olsen, L.E., 2004. Suitability of Asian citrus psyllid, Diaphorina citri,as or species of parasitoids better suited to Florida conditions, aug- prey for ladybeetles. BioControl 49, 417–431. Naranjo, S.E., Ellsworth, P.C., 2005. Mortality dynamics and population regulation in mentative release of T. radiata especially in spring, and possibly Bemisia tabaci. Entomologia Experimentalis et Applicata 116, 93–108. application of entomopathogens such as Paeciliomyces fumarosa Persad, A., Hoy, M.A., 2004. Predation by Solenopsis invicta and Blatella asahinai on and Beauvaria bassiana (Hall, 2008; Qureshi et al., 2009). Sustain- Toxoptera citricida parasitized by Lysiphlebus testaceipes and Lipolexis oregmae on citrus in Florida. Biological Control 30, 531–537. able management of psyllid and citrus greening will require an Pfannenstiel, R.S., Booth, W., Vargo, E.L., Schal, C., 2008. The Asian cockroach integrated strategy that includes biotic mortality components. Blattella asahinai Mizukubo (Dictyoptera: Blattellidae): a new predator of lepidopteran eggs in south Texas soybean. Annals of the ESA 101, 763–768. Qureshi, J.A., Stansly, P.A., 2007. Integrated approaches for managing the Asian Acknowledgments citrus psyllid Diaphorina citri (Homoptera: Psyllidae) in Florida. In: Proceedings Florida State Horticultural Society 3–4 June 2007. Palm Beach, FL, pp. 110–115. We are thankful to Florida Citrus Production Research Advisory Qureshi, J.A., Rogers, M.E., Hall, D.G., Stansly, P.A., 2009. Incidence of invasive Diaphorina citri (Hemiptera: Psyllidae) and its introduced parasitoid Tamarixia Council (FCPRAC) for funding, and our anonymous reviewers, and radiata (Hymenoptera: Eulophidae) in Florida citrus. Journal of Economic M. Triana for technical assistance. Entomology 102, 247–256. Qureshi, J.A., Stansly, P.A., 2008. Rate, placement, and timing of aldicarb applications to control Asian citrus psyllid, Diaphorina citri Kuwayama (Hemiptera: References Psyllidae), in oranges. Pest Management Science 64, 1159–1169. Rogers, M.E., 2008. General pest management considerations. Citrus Industry 89, Aubert, B., Grisoni, M., Villemin, M., Rossolin, G., 1996. A case study of 12–17. huanglongbing (greening) control in Réunion. In: da Graça, J.V., Moreno, P., Rogers, M.E., Stansly, P.A., Stelinski, L.L., 2008. Florida citrus pest management Yokomi, R.K. (Eds.), Proceedings, 13th Conference of the International guide: Asian citrus psyllid and citrus leafminer. Entomol. Nematol. Dept., Fla. Organization of Citrus Virologists (IOCV). University of California Press, Coop. Ext. Serv., Inst. Food Agri. Sci., Univer. Fla., ENY-734. Available from: Riverside, CA, pp. 276–278. http://edis.ifas.ufl.edu/IN686. 136 J.A. Qureshi, P.A. Stansly / Biological Control 50 (2009) 129–136

Roistacher, C.N., 1996. The economics of living with citrus diseases: huanglongbing Stansly, P.A., Qureshi, J.A., 2008. Controlling Asian citrus psyllid: sparing biological (greening) in Thailand. In: da Graça, J.V., Moreno, P., Yokomi, R.K. (Eds.), control. Citrus Industry 89, 20–24. Proceedings, 13th Conference of the International Organization of Citrus Stansly, P.A., Liu, T.X., Schuster, D.J., 2002. Effects of horticultural mineral oils on a Virologists (IOCV). University of California Press, Riverside, CA, pp. 279–285. polyphagous whitefly, its plant hosts and its natural enemies. In: Beattie, G. SAS Institute, 2004. SAS for Windows, Version 9.1. Cary, NC, USA. et al. (Eds.), Spray Oils Beyond 2000. University of Western Sydney Press, pp. Shapiro, S.S., Wilk, M.B., 1965. An analysis of variance test for normality. Biometrika 120–133. 52, 591–611. Tsai, J.H., Wang, J.J., Liu, Y.H., 2002. Seasonal abundance of the Asian citrus psyllid, Shapiro, S.S., Wilk, W.B., Chen, H.J., 1968. A comparative study of various tests of Diaphorina citri (Homoptera: Psyllidae) in southern Florida. Florida normality. Journal of the American Statistical Association 63, 1343–1372. Entomologist 85, 446–451. Shivankar, V.J., Rao, C.N., Singh, S., 2000. Studies on citrus psylla, Diaphorina citri Van den Berg, M.A., Dippenaar-Shoeman, A.S., Deacon, V.E., Anderson, S.H., 1992. Kuwayama: a review. Agricultural Reviews 21, 199–204. Interactions between citrus psylla, Trioza erytreae (Hemiptera: Triozidae) and Smith, H.S., DeBach, P., 1942. The measurement of the effect of entomophagous spiders in an unsprayed citrus orchard in the Transvaal Lowveld. Entomophaga insects on population densities of their hosts. Journal of Economic Entomology 37, 599–608. 35, 845–849. Van Driesche, R.G., Elkinton, J.S., Bellows, T.S., 1994. Potential use of life tables to Snoddy, E.T., Appel, A.G., 2008. Distribution of Blattella asahinai (Dictyoptera: evaluate the impact of parasitism on population growth of Phyllonorycter Blattellidae) in Southern Alabama and Georgia. Annals of the Entomological crataegella (Lepidoptera: Gracillariidae). In: Maier, C. (Ed.), Integrated Society of America 101, 397–401. Management of Tentiform Leafminers, Phyllonorycter (Lepidoptera: Southwood, T.R.E., 1978. The construction, description and analysis of age-specific Gracillariidae) spp., in Leafminers in Apple Orcbards. Thomas Say Publications life tables. In: Ecological Methods with Particular Reference to the Study of in Entomology, Entomological Society of America, Lanham, MD, USA. Insect Population, second ed. Chapman & Hall, London. Wenninger, E., Hall, D.G., 2007. Daily timing of mating and age at reproductive SPSS, 2004. SPSS for Windows, Version 13.0. SPSS Inc., Chicago, IL, USA. maturity in Diaphorina citri (Hemiptera: Psyllidae). Florida Entomologist 90, Stansly, P.A., Schuster, D.J., Liu, T.X., 1997. Apparent parasitism of Bemisia argentifolii 715–722. (Homoptera: Aleyrodidae) by Aphelinidae (Hymenoptera) on vegetable crops Xiao, Y., Qureshi, J.A., Stansly, P.A., 2007. Contribution of predation and parasitism and associated weeds in south Florida. Biological Control 9, 49–57. to mortality of citrus leafminer Phyllocnistis citrella Stainton (Lepidoptera: Stansly, P.A., Conner, J.M., Pomerinke, M.A., 1999. Insecticidal control of Asian Gracillariidae) populations in Florida. Biological Control 40, 396–404. cockroach in orange groves, 1998. Management Test 24, 93–94.